Audio: ASRC: Optimize for HiFi5 function asrc_fir_filter32()

This change improves efficiency of FIR filter compute. The
FIR calculation is unrolled by four. The coefficients and
delay line reads are changed to 128 bits 4x int32_t reads.

The MAC instruction is changed from dual-MAC to quad-MAC.
The FIR accuracy improves a bit due to internal Q17.47 format
instead of Q1.31.

The saving is 1.7 MCPS, from 27.2 to 25.5 MCPS with 32 bit
44.1 to 48.8 kHz stereo push mode ASRC.

Signed-off-by: Seppo Ingalsuo <seppo.ingalsuo@linux.intel.com>

Author: singalsu

src/audio/asrc/asrc_farrow_hifi5.c

@@ -100,11 +100,16 @@ void asrc_fir_filter16(struct asrc_farrow *src_obj, int16_t **output_buffers,
 void asrc_fir_filter32(struct asrc_farrow *src_obj, int32_t **output_buffers,
 		       int index_output_frame)
 {
-	ae_f32x2 prod;
-	ae_f32x2 buffer01 = AE_ZERO32(); /* Note: Init is not needed */
-	ae_f32x2 filter01 = AE_ZERO32(); /* Note: Init is not needed */
-	ae_f32x2 *filter_p;
-	ae_f32x2 *buffer_p;
+	ae_valignx2 align_filter;
+	ae_valignx2 align_buffer;
+	ae_f64 prod0;
+	ae_f64 prod1;
+	ae_f32x2 buffer23;
+	ae_f32x2 buffer01;
+	ae_f32x2 filter01;
+	ae_f32x2 filter23;
+	const ae_int32x4 *filter_p;
+	const ae_int32x4 *buffer_p;
 	int n_limit;
 	int ch;
 	int n;
@@ -115,7 +120,7 @@ void asrc_fir_filter32(struct asrc_farrow *src_obj, int32_t **output_buffers,
 	 * 'n_limit' is therefore stored to reduce redundant
 	 * calculations. Also handle possible interleaved output.
 	 */
-	n_limit = src_obj->filter_length >> 1;
+	n_limit = src_obj->filter_length >> 2;
 	if (src_obj->output_format == ASRC_IOF_INTERLEAVED)
 		i = src_obj->num_channels * index_output_frame;
 	else
@@ -124,55 +129,41 @@ void asrc_fir_filter32(struct asrc_farrow *src_obj, int32_t **output_buffers,
 	/* Iterate over each channel */
 	for (ch = 0; ch < src_obj->num_channels; ch++) {
 		/* Pointer to the beginning of the impulse response */
-		filter_p = (ae_f32x2 *)&src_obj->impulse_response[0];
+		filter_p = (ae_int32x4 *)&src_obj->impulse_response[0];
 
 		/* Pointer to the buffered input data */
 		buffer_p =
-			(ae_f32x2 *)&src_obj->ring_buffers32[ch]
+			(ae_int32x4 *)&src_obj->ring_buffers32[ch]
 			[src_obj->buffer_write_position];
 
 		/* Allows unaligned load of 64 bit per cycle */
-		ae_valign align_filter = AE_LA64_PP(filter_p);
-		ae_valign align_buffer = AE_LA64_PP(buffer_p);
+		align_filter = AE_LA128_PP(filter_p);
+		align_buffer = AE_LA128_PP(buffer_p);
 
-		/* Initialise the accumulator */
-		prod = AE_ZERO32();
+		/* Initialise the accumulators */
+		prod0 = AE_ZERO64();
+		prod1 = AE_ZERO64();
 
 		/* Iterate over the filter bins */
 		for (n = 0; n < n_limit; n++) {
-			/* Read two buffered samples at once */
-			AE_LA32X2_IP(buffer01, align_buffer, buffer_p);
+			/* Read four buffered samples */
+			AE_LA32X2X2_IP(buffer01, buffer23, align_buffer, buffer_p);
 
-			/* Store two bins of the impulse response */
-			AE_LA32X2_IP(filter01, align_filter, filter_p);
+			/* Load four coefficients of the impulse response */
+			AE_LA32X2X2_IP(filter01, filter23, align_filter, filter_p);
 
 			/* Multiply and accumulate */
-			AE_MULAFP32X2RS(prod, buffer01, filter01);
+			AE_MULAAF2D32RA_HH_LL(prod0, prod1, buffer01, buffer23, filter01, filter23);
 		}
 
-		/* Shift left after accumulation, because interim
-		 * results might saturate during filtering prod = prod
-		 * << 1; will shift after last addition
-		 */
-
-		/* swap LL and HH reusing filter01 to perform
-		 * saturated addition of both halves
-		 */
-		filter01 = AE_SEL32_LH(prod, prod);
-
-		/* Add up the lower and upper 32 bit data of the
-		 * 'prod' prod = AE_ADD32_HL_LH(prod, prod); fix using
-		 * saturated addition
-		 */
-		prod = AE_ADD32S(prod, filter01);
+		/* Add up the two accumulators */
+		prod0 = AE_ADD64S(prod0, prod1);
 
 		/* Shift with saturation */
-		prod = AE_SLAI32S(prod, 1);
+		buffer01 = AE_SLAI32S(AE_ROUND32F48SASYM(prod0), 1);
 
-		/* Store 'prod' in (de-)interleaved format in the output
-		 * buffers
-		 */
-		AE_S32_L_X(prod, (ae_f32 *)&output_buffers[ch][i], 0);
+		/* Store 'buffer01' in (de-)interleaved format in the output buffers */
+		AE_S32_L_X(buffer01, (ae_f32 *)&output_buffers[ch][i], 0);
 	}
 }